Gaps in Coverage for Pregnant Women and New Mothers in HUSKY A (Medicaid)

Transcription

1 Gaps in Coverage for Pregnant Women and New Mothers in HUSKY A (Medicaid) Mary Alice Lee, PhD Sarah Esty October 2012 [Type a quote from the document or the summary of an interesting point. You can position the text box anywhere in the document. Use the Drawing Tools tab to change the formatting of the pull quote text box.] Connecticut Voices for Children 0

2 Gaps in Coverage for Pregnant Women and New Mothers in HUSKY A (Medicaid) October 2012 KEY FINDINGS KEY FINDINGS There is growing recognition of the importance of good maternal health before pregnancy, including the time between the birth of one child and the conception of the next. This period offers a window of opportunity for addressing family planning and any outstanding health issues, with the goal of improving maternal health for subsequent pregnancies. Access to care is critically important during this period. We investigated coverage continuity in the prenatal and postpartum periods for women who gave birth while enrolled in HUSKY A (Medicaid managed care) in Findings: Prior to the birth, pregnant women were enrolled in HUSKY A an average of 6 months, with 42 percent continuously enrolled for nine months. Following the birth, over 70 percent of new mothers were continuously enrolled for nine months. The likelihood of gaps or loss of coverage in the postpartum period was far greater for mothers who qualified for coverage during pregnancy through 60 days postpartum, compared with mothers in ongoing Medicaid family and child coverage groups. At nine months after the birth more than half of all mothers with pregnancy-related coverage had experienced a gap or lost coverage. Coverage continuity also varied by where the mother lived, suggesting that community-based application support and administrative procedures in the Medicaid agency s district offices play a role in facilitating coverage in the postpartum period. These findings have implications for Medicaid program administration and for design of an integrated eligibility management system that will facilitate seamless automated coverage transitions between Medicaid and other insurance options offered by Connecticut s Health Insurance Exchange in INTRODUCTION The prenatal period is widely recognized as an exceptional time for promoting maternal health and optimal birth outcomes. Regular prenatal screenings can identify problems early, allowing for timely treatment that can significantly reduce risks and improve outcomes. For women with existing risk factors, such as obesity, diabetes, and hypertension, active monitoring and treatment to control the conditions during the prenatal period are especially important to ensure that women and their babies are as healthy as possible. Furthermore, during prenatal visits, health professionals can provide important information and support for good nutrition, reduction of risky behaviors like smoking or drinking, and other ways to promote a healthy pregnancy. As clinicians and policy makers seek to improve birth and maternal outcomes beyond those achievable through prenatal care, a growing focus for research and policy development has been on the importance of maternal health before the prenatal period in the pre-conception period--and across a woman s lifetime. 1 1 Board on Population Health and Public Health Practice, Institute of Medicine. Clinical preventive services for women: closing the gaps. Washington, DC: National Academies Press, Available for download at: Connecticut Voices for Children 1

3 A subset of this work centers on the interconceptional or interpregnancy period: the time between the birth of one child and the conception of the next. 2 This period poses special challenges for new mothers who may be in need of extra health care as they recover from giving birth and adapt to parenthood. It also offers a window of opportunity to address family planning and any outstanding health issues, with the goal of improving maternal health for a subsequent pregnancy. Researchers are also increasingly looking at access and quality of care for improving health outcomes. Continuity of health insurance coverage is at the heart of both, as access to care largely depends on health insurance and quality of care depends on delivering the right services at the right time. 3 Discontinuous coverage limits access, disrupts ongoing care, and is costly for the patient and the provider. Furthermore, coverage gaps in publicly funded programs generate significant costs for states that must process additional enrollment and disenrollment paperwork, and higher costs for the care when patients are re-enrolled. 4 A number of factors contribute to both enrollment and continuity of publicly-funded coverage. This report examined data from Connecticut HUSKY Program for the purpose of identifying factors associated with continuous health insurance coverage for women who gave birth in 2009 while on HUSKY A (Medicaid). We found that mothers who qualified for coverage by virtue of becoming pregnant were more likely to lose coverage following the birth, compared with mothers who were in other coverage groups when they gave birth. Coverage continuity also varied by where the mother lived, suggesting that administrative procedures in the Medicaid agency s district offices and the availability of community-based application support may affect coverage continuity in the postpartum period. HUSKY Program Eligibility in Pregnancy In 2009, pregnant women in Connecticut were eligible for Medicaid coverage during pregnancy and for 60 days postpartum if they lived in households with family income less than 250 percent of the federal poverty level (FPL). 5 Legal immigrant women were covered, including those who had been in the US less than 5 years. 6 During pregnancy, women were covered for pregnancy-related care and the full range of benefits available to other Medicaid beneficiaries. 7 In 2009, pregnant women were enrolled in HUSKY A (Medicaid managed care), with statewide mandatory enrollment for children, parents, and pregnant women who qualified for Medicaid. The only exceptions to managed care enrollment for pregnant women were for 2 See, for example: Lu M, Kotelchuck M, Culhane J, Hobel C, Klerman L, Thorp J. Preconception care between pregnancies: the content of internatal care. Maternal Child Health Journal, 2006; 10:S107-S Kenney G, Pelletier J. Monitoring duration of coverage in Medicaid and CHIP to assess program performance and quality. Academic Pediatrics 2011, 11( 3S): S34-S41. 4 In California in , the administrative costs of re-enrolling children in the state s Medicaid program after gaps in coverage were estimated at approximately $140-$160 per child. Costs in the 6 months following a gap in coverage were higher than for those without gaps, and increased further for those with longer gaps. (Fairbrother G, Schuchter J. Stability and churning in Medi-Cal and Healthy Families. The California Endowment, March Available at: b7824db1cec3.pdf). 5 In 2009 and 2010, 250% FPL was $36,425 for a family of two. For the purpose of eligibility determination within the pregnancy-based coverage categories, a pregnant woman is counted as two persons; a woman who becomes pregnant while enrolled in family or child coverage is not counted as two. For pregnant teens, parental income counts for eligibility determination unless the teen leaves the household or does not renew coverage. 6 Pregnant women who were legal permanent residents in the US less than 5 years were eligible for state-funded Medicaid coverage in In 2009, after Congress passed the Children s Health Insurance Program Reauthorization Act, Connecticut amended the Medicaid State Plan and effective April 1, began collecting federal matching funds for coverage of pregnant women. In 2011, after the close of this study period, state-funded coverage for non-pregnant legal permanent resident adults who had been in the US less than 5 years was eliminated by the Connecticut General Assembly. Effective July 1, 2011, about 4,900 adults (mostly parents and caretaker relatives of children in the Medicaid program) lost coverage. 7 In 2009 and earlier years, treatment for tobacco dependence was not a covered benefit in Connecticut s Medicaid program. Connecticut Voices for Children 2

4 those women who enrolled in the third trimester of pregnancy and were in care with non-participating prenatal care providers; they were covered in fee-for-service Medicaid. Connecticut Department of Social Services case workers in the district offices and regional processing units assign pregnant women to the coverage groups that are most advantageous for them. For example, a pregnant woman whose entire family needs coverage will be assigned to the family coverage group. A pregnant teen may be assigned to a child group because the coverage is not tied to the time limits inherent in pregnancy coverage. A pregnant woman may be assigned to a pregnancy-related group to take advantage of the higher income threshold and household composition counting rules (a pregnant woman is counted as two persons). Thus, most pregnant women qualify for Medicaid as: Parents of enrolled or eligible children, living in households with income less than 185 percent of the federal poverty level (family coverage 8 ; or Adolescents living in households with income less than 185 percent of the federal poverty level (family coverage and transitional medical assistance); or Adolescents in child-only coverage groups, eligible till age 19 if living in households with income less than 185 percent FPL or up to age 21 if they meet certain very low income and asset guidelines; or Women who are eligible for HUSKY A coverage by virtue of being pregnant and living in households earning less than 250 percent of the federal poverty level (pregnant woman coverage groups, with pregnant woman counted as two persons for the purpose of eligibility determination). Results of an earlier investigation showed that a sizeable proportion of mothers who gave birth in 2001 (43%) was enrolled prior to becoming pregnant, either as parents of enrolled children (family coverage groups) or as adolescents (in family coverage groups or children s coverage group). 9 HUSKY Program Coverage for New Mothers Under federal law, women covered by Medicaid through pregnancy-related coverage groups maintain coverage for 60 days postpartum. 10 After 60 days, women cease to be eligible for the pregnancy-related coverage and receive a notice from the state Department of Social Services informing them that the program has ended. 11 In 2009, they also received a notice informing them that they were disenrolled from managed care. Reportedly, many eligible women experience gaps or lose coverage altogether at this time Family coverage includes transitional medical assistance for families with earnings or child support payments that would otherwise put them over the Medicaid income eligibility level. 9 Unpublished results of study of enrollment and the timing of prenatal care initiation among women who gave birth with HUSKY A coverage in For detailed results, contact Mary Alice Lee at Connecticut Voices for Children. 10 In 2009 and 2010, legal permanent residents who had been in the US for less than 5 years were eligible for state-funded Medicaid coverage. In 2011, after the close of this study period, state-funded coverage for non-pregnant legal permanent resident adults who had been in the US less than 5 years was eliminated by the Connecticut General Assembly. Effective July 1, 2011, about 4,900 adults (mostly parents and caretaker relatives of children in the Medicaid program) lost coverage. 11 Currently, the notice of discontinuance for postpartum women in the M01 coverage group informs them that If you are a nonpregnant female or male (sic) who is in need of family planning services, you may be eligible for the Family Planning Coverage- Limited Benefit.If you would like to apply for the Family Planning Coverage-Limited Benefit, please contact DSS at the telephone number listed above. (Source: 2007 Notice of Discontinuance HUSY A M01 Post-Partum Women, obtained from the Department of Social Services.) These women do not get an automatic renewal form, but regional office staff should contact them by phone or mail to determine whether the mother is eligible for another coverage group. 12 Anecdotal reports from HUSKY Infoline care coordinators and from community-based social services providers who gathered to discuss retention. Connecticut Voices for Children 3

5 Some of these gaps result from women s mistaken belief that they are no longer eligible a misunderstanding due in no small part to the confusing disenrollment notices (see text box below). Once a mother in the pregnancy-related group gives birth, the Department of Social Services Central Processing Unit assigns the baby to the newborn coverage group (F10) and notifies the mother s Medicaid Excerpt from Notice Sent to Women Being Disenrolled from Pregnancy Coverage Groups P02 HUSKY A for Pregnant Women AU XXXXXXX Your Medical assistance will be discontinued on [date]. We are taking action for the following reason(s): YOU DIDN T PROVE YOU ARE PREGNANT, OR HAVEN T BEEN PREGNANT LONG ENOUGH, OR YOU FAILED TO VERIFY YOUR DUE DATE. Policy Reference: , , Source: Notice obtained from the Connecticut Department of Social Services; on file at Connecticut Voices. case worker. 13 Depending on when the regional office receives notice of the baby s birth, the notice may trigger the worker to manually close the pregnancy-related coverage and assign the mother to a postpartum coverage group for 60 days. Notices are automatically sent to the family when the baby is added and when the mother s coverage group is changed. A third notice may be sent to the family when the 60 days postpartum coverage is granted. If notice of the birth is not received and processed in a timely fashion, the eligibility management system automatically sends out a notice and short, prefilled renewal form ten months after pregnancy coverage was granted. [Note: Connecticut Voices was unable to determine whether the notice is accompanied by a pre-filled 4-page HUSKY renewal form or an 8-page redetermination form that requires the new mother to provide information and documentation that would not otherwise be required for ongoing HUSKY coverage.] Mothers who reach the end of the 60-day postpartum coverage should be screened for eligibility in another coverage group, a process that involves determining income eligibility and securing child support cooperation. However, if the redetermination is not initiated, the coverage terminates automatically. Due to the volume of applications and redeterminations, workers are not always able to conduct timely manual reviews of eligibility for postpartum women. [Note: Connecticut Voices was unable to obtain a copy of the discontinuance notice that is sent at the end of the 60 days postpartum.] Pregnant women whose coverage was not related to pregnancy i.e., those who were covered in the family and child coverage groups when they became pregnant--are subject to annual eligibility redetermination, irrespective of their pregnancies. This means that they may come up for redetermination at any time during the pregnancy or might not face redetermination until some months following the birth. PURPOSE OF THIS STUDY To describe coverage continuity for pregnant women and new mothers who gave birth while insured in the HUSKY A Program (Medicaid). 13 Once the new ConnectCT system for managing and providing information on application and renewal status is up and running, notices about births will be sent from the Central Processing Unit to the Regional Processing Units rather than specific case workers for eligibility determination. Connecticut Voices for Children 4

6 To investigate the association between Medicaid coverage group and loss of coverage in the postpartum period. METHODS We conducted a retrospective cohort analysis of enrollment in the months before and following births to mothers with HUSKY A (Medicaid managed care) coverage. Using a survival analysis approach, we determined by coverage group which mothers were continuously covered prior to the birth and postpartum. After adjusting for other factors that may have affected coverage continuity, we determined the likelihood of continuous coverage for the nine months after birth and the difference in average length of continuous coverage by coverage group in which the mothers were enrolled at the time of the births. Data As part of a larger program of independent performance monitoring in the HUSKY Program, we linked 2009 birth data with HUSKY Program enrollment and Medicaid eligibility data provided by the Department of Social Services. 14 Birth data for Connecticut residents were obtained from the Connecticut Department of Public Health. 15 Both datasets contain personal identifiers (social security number, name, maternal date of birth) that were used to link the records according to a matching algorithm that has been used for more than ten years in Connecticut. 16 For each mother whose records matched, monthly HUSKY A (Medicaid managed care) enrollment data for the nine months before and the nine months after the birth month were compiled longitudinally. The nineteen month observation period was chosen in order to capture the entire prenatal period and the early months postpartum, as well as at least one annual redetermination of eligibility when HUSKY enrollees are at greater risk for losing coverage. 17 In each of the months that the mothers were enrolled, the corresponding Medicaid coverage categories were identified in the file. 18 With the exception of the HUSKY coverage category and primary language, all other data that describe maternal characteristics were derived from the birth certificates. Study Sample Based on the data linkage described above, 14,693 babies were born to Connecticut women with publiclyfunded coverage in 2009 (38.7% of all in-state Connecticut births), including 11,995 births to 11,811 mothers with HUSKY A (Medicaid managed care) coverage. For the purposes of this study, mothers who gave birth while enrolled in HUSKY B (CHIP) coverage (28 mothers of 29 babies) and those with Medicaid 14 Independent performance monitoring in the HUSKY Program is state-funded and conducted by Connecticut Voices under a contract between the Department of Social Services and the Hartford Foundation for Public Giving (contract # 064HFP-HUO- 03/10DSS1001ME, 4/1/10-6/30/13). 15 Birth data were released to Connecticut Voices for Children by the Department of Public Health, with the approval of its Human Investigations Committee, for linkage with HUSKY Program enrollment data (Protocol #686, approved March 9, 2011). Copies of the linked dataset were returned to the Departments of Public Health and Social Services in accord with their interagency data-sharing agreement (DPH LOG # ). 16 For a more detailed description of the data matching algorithm, see Births to Mothers with HUSKY Program and Medicaid Coverage: 2009, available at Lee MA, Hero J. HUSKY Program Enrollment Dynamics: Coverage Continuity, Gaps in Coverage and Retention. New Haven CT: Connecticut Voices for Children, April Available at: 18 The Department of Social Services assigns the labels and codes. Connecticut Voices for Children 5

7 fee-for-service coverage (2,630 mothers of 2,669 babies) were not included in the analyses. Mothers in HUSKY A who had multiple births (184) were counted just once and three others were excluded because of inconsistencies in what should have been duplicate records for three twin births. Records for 437 women in HUSKY A were also removed from the analyses because of 1) insufficient data for classifying coverage group at the time of the birth (335 mothers), 2) missing data for personal characteristics used to adjust analyses of coverage continuity (50 missing maternal education, 24 with race/ethnicity unknown), 3) attribution to coverage groups that appear to have been reported in error (2 mothers in Medicaid for newborns), 4) assignment to coverage groups with fewer than ten mothers at the time of birth (20 mothers in coverage groups for children in subsidized adoptions, medically needy families, and Medicaid spenddown), or 5) some combination of the aforementioned reasons. Therefore, the findings are based on the coverage experience of 11,374 mothers or 96.3% of those who gave birth while enrolled in HUSKY A in Variables In order to study the association between coverage category and coverage continuity, we grouped mothers in the following way according to the Medicaid coverage group in which they were enrolled in the month they gave birth: Pregnancy-related coverage groups: For pregnant women who qualified for Medicaid by virtue of being pregnant and living in households with income less than 250% FPL (Medicaid/Pregnant Women--P01; Healthy Start Medicaid--P02); Family coverage groups: For children and parents or relative caregivers with household income less than 185 % FPL (HUSKY A for Families--F07) and those in family groups with household income that exceeded 185% FPL due to earnings (Transitional Medicaid for Families with Earnings--F03) or child support (Child Support Extension--F04); Child coverage groups: For children in families with income less than 185% FPL but whose parents were not covered (HUSKY A for Children F25) and children 19 to 20 in families with much lower income and assets (Medicaid for Children under 21--F12). Multivariate analyses of the association between coverage category and coverage continuity were adjusted for the following factors that might have affected eligibility, enrollment, and continuous enrollment pre-and post-birth: Maternal age (19 and under, 20-29, 30-39, 40 and older); Maternal race/ethnicity (white non-hispanic; black non-hispanic; Hispanic or Latino of any race; and those identifying with other races, including Asian non-hispanic and Indian non-hispanic); Maternal education (less than 12 years education, high school or 12 years education, more than 12 years education); Maternal residence (grouped by DSS regional office catchment area 19 ); 19 DSS regional office catchment areas: NORTHERN, serving the towns of Andover, Avon, Ashford, Berlin, Bloomfield, Bolton, Bristol, Brooklyn, Burlington, Canterbury, Canton, Chaplin, Columbia, Coventry, Eastford, East Granby, East Hartford, East Windsor, Ellington, Enfield, Farmington, Glastonbury, Granby, Hampton, Hartford, Hebron, Killingly, Manchester, Mansfield, Marlborough, New Britain, Newington, Plainfield, Plainville, Plymouth, Pomfret, Putnam, Rocky Hill, Scotland, Simsbury, Somers, Southington, South Windsor, Stafford, Sterling, Suffield, Thompson, Tolland, West Hartford, Union, Vernon, Wethersfield, Willington, Windham, Windsor, Windsor Locks and Woodstock; SOUTHERN, serving the towns of Ansonia, Bethany, Branford, Bozrah, Chester, Clinton, Colchester, Cromwell, Deep River, Derby, Durham, East Haddam, Connecticut Voices for Children 6

8 Primary language (English, non-english); 20 and Birth order (first birth, second or higher birth). Analytic Approach We conducted essentially two simple survival analyses of coverage, one for the nine months prior to the birth month and one for the nine months following the birth month. Before the birth, women were added to the analysis of coverage when they were first enrolled in the HUSKY program and were subsequently continuously enrolled through the birth month. 21 Post-birth, women were included in the analyses for the months of continuous coverage and dropped from the analyses if they experienced gaps of a month or more or lost coverage in the nine months following the birth month. While the women were grouped by coverage group in the month of birth, coverage continuity was determined by identifying periods of uninterrupted coverage in any coverage group before or after the birth. 22 The percentages of women in each coverage grouping who were covered (and subsequently continuously enrolled) were calculated by month for the nine months prior to the birth. The percentages of new mothers in each coverage grouping who were covered each month postpartum (continuously since the birth month) were calculated. Pre-and post-birth, the average number of months of enrollment were also calculated for periods of continuous coverage. Significant associations between coverage group and maternal characteristics were investigated with chi-square tests. The likelihood of being continuously enrolled in the nine months postpartum, depending on coverage category, was determined using a logistic regression model adjusted for maternal age, race/ethnicity, primary language, residence, education, and parity. Reference levels for the covariates were arbitrarily assigned to the largest subgroups. The number of months of continuous post-birth coverage was determined using a multivariate linear regression model that took into account coverage category when controlling for other variables. Significant associations between the dependent and independent variables were determined using p<0.05. For data management, we used ACCESS and EXCEL ; for data analyses, we used R and SPSS. East Hampton, East Haven, East Lyme, Essex, Franklin, Griswold, Groton, Guilford, Haddam, Hamden, Killingworth, Lebanon, Ledyard, Lisbon, Lyme, Madison, Meriden, Middlefield, Middletown, Milford, Montville, New Haven, New London, North Branford, North Haven, North Stonington, Norwich, Old Lyme, Old Saybrook, Orange, Portland, Preston, Salem, Seymour, Shelton, Sprague, Stonington, Voluntown, Wallingford, Waterford, Westbrook, West Haven and Woodbridge; and WESTERN, serving the towns of Barkhamsted, Beacon Falls, Bethel, Bethlehem, Bridgeport, Bridgewater, Brookfield, Canaan, Cheshire, Colebrook, Cornwall, Danbury, Easton, Fairfield, Goshen, Hartland, Harwinton, Kent, Litchfield, Middlebury, Monroe, Morris, Naugatuck, New Fairfield, New Hartford, New Milford, Newtown, Norfolk, Norwalk, North Canaan, Oxford, Prospect, Redding, Ridgefield, Roxbury, Salisbury, Sharon, Sherman, Southbury, Stratford, Thomaston, Torrington, Trumbull, Warren, Washington, Waterbury, Watertown, Weston, Westport, Winchester, Wolcott and Woodbury. 20 Data on primary language from HUSKY enrollment records. 21 For example, a woman who had coverage 9 and 8 months before the birth, lost coverage in the 7 th month before the birth, and resumed coverage in the 6 th month before the birth through the birth would be considered to have 6 months of continuous prebirth coverage. 22 For example, a woman who was enrolled in a pregnancy-related coverage category (P01, P02) in the month she gave birth would be considered continuously enrolled in the 9 months postpartum if she was enrolled in postpartum coverage (M01, M02) or in a family coverage group (F03, F04, F07) without a gap in coverage. Connecticut Voices for Children 7

9 Limitations Interpretation of the findings and conclusions based on these findings are limited by several methodological considerations: The findings are based on secondary analyses of enrollment data collected by the Connecticut Department of Social Services for administering a capitated statewide Medicaid managed care program with mandatory enrollment. Independent validation of enrollment records was beyond the scope of this investigation. Fee-for-service Medicaid data were not available for determining whether what appeared to be gaps or loss of coverage were in fact closed with temporary fee-for-service coverage granted in advance of health plan selection or reinstated after a lapse. Information was not available for determining why individual women enrolled when they did during pregnancy. Information for determining why new mothers lost coverage was not available. Income data were unavailable, so it was not possible to determine which mothers who were eligible during pregnancy were over-income for family coverage in the postpartum period (household income between 185% and 250% FPL). 23 Despite these limitations, the analyses are useful for indicating areas for program improvement and further investigation of HUSKY Program enrollment for pregnant women and new mothers. RESULTS In 2009, there were 11,811 mothers who gave birth while covered by HUSKY A. These analyses are based on the enrollment experience of the majority of those mothers (11,374 or 96.3%). The distribution of mothers by coverage group and maternal characteristics is shown in Table 1. There were significant differences between mothers in different coverage group at the time of birth. Mothers in the pregnancy-related coverage group were younger, more likely to be White non-hispanic, and better educated than mothers in the family coverage groups; they were more likely to be having their first child. Compared with those in the child coverage groups, mothers in the pregnancy-related coverage group were older, White non-hispanic, and better educated; they were less likely to be having their first child. Whereas residence was distributed fairly evenly across the three DSS regions for mothers in the pregnancyrelated and family coverage groups, disproportionately more mothers in the child coverage groups lived in the northern region of the state. 24 Enrollment Prior to the Birth Pregnant women who gave birth in 2009 were continuously enrolled in HUSKY A for an average of 6.1 months prior to the birth (Table 2). Mothers in the pregnancy-related groups were enrolled on average just 4.3 months, compared with 6.9 months for those in child coverage groups and 7.7 months for those in family groups when they gave birth. About half of all mothers with pregnancy-related coverage were 23 Those who were in family coverage (F07) whose household income increased due to earnings should be eligible in transitional Medicaid (F03) for one year. Likewise, if household income increased due to child support payments, the coverage is extended for one year. If this transition between coverage groups occurred without a gap in coverage, these mothers would be counted as continuously eligible. Transitional Medicaid is not available to persons in any other Medicaid coverage group. 24 Department of Social Service regional office catchment area for the Northern region includes offices in New Britain, Hartford, Manchester, and Willimantic. Connecticut Voices for Children 8

10 continuously enrolled beginning in the second trimester or about five months prior to the birth (Figure 1 on following page). Overall, 42 percent of mothers were continuously enrolled for nine months prior to the birth. As might be expected because of the tie between pregnancy and eligibility, those in the pregnancy-related groups were far less likely to have been enrolled continuously in the nine months before the birth (8.5%), compared with mothers in the family (73.3%) and child (57.2%) coverage groups whose eligibility was based on other family circumstances. Figure 1. Continuous Coverage Prior to the Birth by Coverage Group 100% Percent Continuously Enrolled From that Month Through Birth 80% 60% 40% 20% Child Family Pregnancy Related 0% Birth Note: Mothers are grouped by coverage group they were in the month they gave birth. Those who were continuously enrolled for 9 months prior to the birth may have been in other coverage groups but did not experience a gap in coverage before the birth. Enrollment Following the Birth Months Before the Birth Overall, about 71 percent of mothers who gave birth in 2009 were continuously enrolled for nine months following the birth (Table 3). Those in the pregnancy-related groups were continuously enrolled an average of 6.3 months following the birth, compared with 8.3 months for those in family groups and 7.6 months for those in child groups when they gave birth. Mothers in pregnancy-related coverage groups were far less likely than family or child group mothers to have been enrolled continuously in the nine months after pregnancy (54.3%, compared with 86.2% and 70.3% enrolled, respectively). Mothers who were in pregnancy-related groups when they gave birth began to lose coverage beginning at about two months postpartum (Figure 2 on following page). This timing coincides with the end of the federally-defined eligibility period (60 days) associated with Medicaid coverage during pregnancy. At the end of this eligibility period, federal law requires screening for eligibility in other coverage groups prior to disenrollment from coverage. However, by nine months postpartum, more than half of all mothers with pregnancy-related coverage had experienced a gap or lost coverage in HUSKY A. Connecticut Voices for Children 9

11 Figure 2. Continuous Coverage After the Birth by Coverage Group 100% Percent Continuously Enrolled From Birth Through that Month 80% 60% 40% 20% Child Family Pregnancy Related 0% Birth Months After the Birth Note: Mothers are grouped by coverage group they were in the month they gave birth. Those who were continuously enrolled for 9 months postpartum may have been in other coverage groups but did not experience a gap in coverage after the birth. Otherwise, maternal age, education, primary language, and number of prior children (parity) had no effect on the likelihood of being enrolled postpartum. Risk of Losing Coverage in the Postpartum Period After adjusting for other factors that may have affected eligibility and enrollment in the postpartum period, new mothers in the pregnancy-related coverage groups were far more likely to have experienced gaps or loss of coverage in the nine months following the birth (Table 4). In fact, women in the pregnancy-related coverage groups were over five times more likely than those in the family coverage groups and two times more likely than those in the child coverage groups to have lost coverage in the nine months following the birth. Hispanic ethnicity decreased the risk of coverage loss, compared to coverage for White women. Coverage group had a significant effect on the number of months postpartum that new mothers were continuously covered in the HUSKY Program (Table 5). After adjusting for other factors, mothers who were in the pregnancy-related coverage groups at the time of the birth had an average of one to two fewer months of continuous coverage than mothers in the family or child coverage groups. Hispanic mothers were covered slightly longer than White mothers. Where the mother lived, i.e., which Department of Social Services district office managed her eligibility redetermination, was a significant factor in continuity of postpartum coverage. Compared with the New Haven office (reference level), mothers were at greater risk for losing coverage postpartum if they lived in the areas served by the Bridgeport, Danbury, or Hartford offices. Their coverage was on average 1 to 3 weeks shorter if they lived in towns served by the Bridgeport, Danbury, Hartford, Manchester and Stamford offices. The availability of community-based application assistance may also have been a factor. By nine months after the birth, 46 percent of those who were in pregnancy-related coverage had experienced gaps or lost coverage (Table 6). Among those who lost coverage at some point postpartum, about 24 percent were enrolled at nine months, suggesting that they may have been eligible during the Connecticut Voices for Children 10

12 coverage gap. Nearly 1,900 new mothers (34.9% of those with pregnancy-related coverage) were not enrolled ( Figure 3). Figure 3. New Mothers Coverage Status After Giving Birth DISCUSSION The time after a birth provides an opportunity to improve not only the health of new mothers and their babies, but also health and birth outcomes in subsequent pregnancies. Access to physical, oral, and mental health services during the postpartum period facilitates follow-up of conditions discovered during pregnancy, as well as treatment for emerging and ongoing health conditions such as postpartum depression, obesity, diabetes, tobacco and alcohol use, substance abuse, and periodontal disease. Interconceptional care provides opportunities for helping new mothers plan for subsequent pregnancies when they and their families are ready. Both adult and pediatric care providers can play a role in risk-screening and counseling, supporting family planning decisions, and promoting family health. 25 The results of this study show that a significant proportion of women who had Medicaid coverage for the births of their babies lost coverage in the postpartum period, especially women whose coverage was tied to the pregnancy and discontinued at 60 days postpartum. Without data to track disenrollment reasons, it is not possible for us to determine whether women who lost coverage were in fact eligible or were ineligible because they were over-income, moved out of state, voluntarily dropped coverage, relinquished custody of the baby, or failed to cooperate with child support requirements. It is also not possible to determine 25 Cheng TL, Kotelchuck M, Guyer B. Preconception women s health and pediatrics: an opportunity to address infant mortality and family health. Academic Pediatrics 2012, 12(5): Connecticut Voices for Children 11

13 whether new mothers who lost coverage failed to complete the renewal/redetermination process. At the very least, the notice itself and the procedures for renewing coverage may be confusing and unnecessarily daunting. It is also possible that coverage was inadvertently terminated by the Department because of the antiquated eligibility management system and the cumbersome manual procedures for redetermination when eligibility ends. Federal law requires screening for eligibility in other Medicaid coverage groups before termination of pregnancy-related coverage. Recently the Commonwealth Fund, in partnership with the Centers for Disease Control and Prevention, worked with seven states (California, Florida, Illinois, Louisiana, North Carolina, Oklahoma and Texas) to examine Medicaid strategies for improving reproductive health and birth outcomes. 26 These states reported that Medicaid covered between 46 percent (Illinois, 2008) and 70 percent (Louisiana, 2008) of all births (compared with 39% in Connecticut in 2009). Four of the five states that had data reported loss of coverage postpartum at rates ranging from 56 percent (Florida) to 73 percent (California, Louisiana). Coverage continuity in one state was notably different: Illinois reported that just 5 percent of mothers lost coverage at 60 days postpartum. Closer examination of state-specific income eligibility levels shows that in Illinois, the income eligibility level for pregnant women (200% FPL) is closely aligned with the income eligibility level for parents (191% FPL). The other six states have far lower income eligibility levels for parents, meaning that women who have been covered during pregnancy are quite likely to be over-income and lose coverage at 60 days postpartum. Connecticut has taken the following steps to improve and monitor women s reproductive and preconception health for those who rely on the HUSKY Program: Raised income eligibility levels: Income eligibility levels for pregnant women and parents are both relatively high, compared with other states outside New England, and nearly aligned, thus ensuring that many if not most women who qualify during pregnancy will also qualify for coverage once they become parents. Income eligibility levels were raised in 2008 for pregnant women to ensure access to prenatal care and in 2007 to reduce uninsured rates for children whose parents are also eligible for coverage. Expanded Medicaid to cover family planning: Effective March 1, 2012, family planning and related services became available for women and men (including minors) with income up to 250 percent of the federal poverty level who would not otherwise qualify for Medicaid. 27 Covered services include comprehensive physical examinations, screening and treatment of sexually transmitted diseases, voluntary sterilizations, contraceptive services and supplies, HPV vaccinations, and pregnancy tests, when delivered as part of a visit for the purpose of preventing pregnancy or regulating the number and spacing of children. Covered services do not include some interventions that have been shown to be effective in improving maternal health and birth outcomes, such as rubella vaccination, nutrition counseling, treatment of tobacco dependence, and management of diabetes and other chronic diseases prior to pregnancy. Invested in monitoring health in pregnancy and program performance: Connecticut uses Medicaid data, linked Medicaid-vital records data, and public health survey data to identify gaps in services, monitor outcomes, and drive decision-making. Annually since 2000, Connecticut Voices has obtained birth data from the Department of Public Health and linked birth records to HUSKY enrollment records, thus making it possible to identify births to mothers with publicly funded 26 Johnson K. Addressing women s health needs and improving birth outcomes: results from a peer-to-peer state Medicaid learning project (pub.1620 vol.21). New York NY: Commonwealth Fund, August ns_htl_needs_improving_birth_ib.pdf 27 Connecticut Department of Social Services. New family planning coverage groups: Family planning services limited benefit (PB ). Issued March Available at: Connecticut Voices for Children 12

14 coverage. 28 The linked dataset is provided to both the Department of Social Services and the Department of Public Health for further analyses. Additionally, Connecticut Voices has conducted special studies of maternal health care (emergency care, behavioral health services, and oral health care) that have identified gaps in services and areas for improvement. 29 The Department of Public Health reports that Connecticut will participate in the Center for Disease Control s nationwide Pregnancy Risk Assessment Monitoring System (PRAMS). PRAMS is a survey of new mothers that produces data for tracking maternal health before and during pregnancy, with analyses by payer source. Connecticut PRAMS data will be available for analysis in Committed to improving the quality of prenatal and postpartum care: Recently, the Department of Social Services has been working with its Medicaid administrative services contractor, the Department of Public Health, and others to develop an incentive program aimed at rewarding maternity care providers for high quality prenatal and postpartum care. Additionally, the Department s support for development of person-centered medical homes holds promise for improving women s health and health care across the life cycle. Beginning in 2014, the Affordable Care Act will provide many opportunities for improving maternal health and birth outcomes, beginning with difficulties women currently have obtaining continuous, affordable health insurance coverage. Additionally, the Affordable Care Act will help to ensure that all low income women have access to family planning and all other needed health services throughout the life cycle. Medicaid will expand to cover all adults with income up 133 percent of the federal poverty level. Affordable options for individual and small group coverage will be available through health insurance exchanges. Preventive services, including contraceptive care, will be available to all women at no cost. Maternity care will be covered in all policies that become available through the health insurance exchanges. New care arrangements, such as person-centered medical homes with incentives to improve health care quality, will help to ensure early detection and timely treatment for conditions associated with poor maternal health and birth outcomes. Seamless transitions between programs and health insurance products will be facilitated by integrated state-of-the-art eligibility management systems that ensure coverage coordination between Medicaid and health insurance exchange options. Taken all together, these aspects of health reform will go a long way towards improving coverage continuity for Connecticut s women and their new families. CONCLUSION AND RECOMMENDATIONS Coverage continuity in the postpartum period is key to ensuring access to needed health services. Just at the time when many new mothers in low income families are recovering from childbirth, adjusting to new parenthood, and perhaps overwhelmed with new responsibilities, they are also at risk for losing health insurance coverage. The fact that so many women lose coverage postpartum, as this study shows, suggests lost opportunities for taking advantage of policies designed to promote family health and access to care. The Department of Social Services, working with its community-based partners and policymakers, can take the following steps to facilitate renewal and improve coverage continuity for new mothers: 28 See Lee MA, Siegel J, Learned A. Births to mothers with HUSKY Program and Medicaid coverage: New Haven, CT: Connecticut Voices for Children, December Available at: 29 See the following reports from Connecticut Voices for Children: Behavioral Health Care for New Mothers in HUSKY A, July Emergency Care for New Mothers and Pregnant Women in HUSKY A, January Dental Care for New Mothers in HUSKY A, October Connecticut Voices for Children 13

15 Conduct a systematic review of coverage continuity in the postpartum period to determine the leading reasons for gaps and loss of coverage and to identify remedies for addressing procedural barriers, including: o Revise notices that go to pregnant women and new mothers; o Send pre-filled 4-page HUSKY applications to all new mothers who must renew coverage; o Adopt procedures that prioritize renewal processing for new mothers, just as applications for pregnant women are expedited; o Do not issue automatic discontinuances without a thorough review of eligibility in other coverage categories, including Medicaid family planning limited benefit coverage if over-income for HUSKY A family or child coverage; o Ensure that the Medicaid case workers and the HUSKY application center inform mothers who are over-income for HUSKY A about the other state-supported coverage options (Medicaid family planning coverage, Charter Oak Health Plan) and urge them to contact 2-1-1/HUSKY Infoline for help obtaining coverage before coverage ends; Alert community-based social services and clinical care providers, including pediatricians, to problems with postpartum coverage continuity and urge them to reach out to new mothers to help prevent gaps in coverage; and Expand person-centered medical home and maternity care pay-for-performance initiatives to include measures aimed at measuring and rewarding high quality well-woman care provided throughout the life cycle by women s health care specialists and other primary care providers. Implementation of the Affordable Care Act provides an opportunity for creating an integrated eligibility management system that will facilitate seamless automated coverage transitions between Medicaid and other insurance options from the Health Insurance Exchange. Advocates and policy makers should pay particular attention to simplifying and expediting coverage transitions for new families. ACKNOWLEDGEMENTS This report was prepared by Connecticut Voices for Children under a contract between the Department of Social Services and the Hartford Foundation for Public Giving, with a grant from the Hartford Foundation to Connecticut Voices for Children. This report was prepared by Mary Alice Lee, Ph.D., Senior Policy Fellow, and Sarah Esty, Policy Fellow. Amanda Learned of MAXIMUS Health Services, Inc. prepared the dataset that was analyzed by Sarah Esty. Rachel Chen assisted in the early stages of this project while she was a student at the Yale School of Public Health. Sharon Langer, J.D., Senior Policy Fellow, and Kathy Misset, Consultant and Project Manager for the Department of Social Services Organizational and Skill Development and the University of Connecticut School of Social Work, critiqued the report and provided invaluable background information. This publication does not express the views of the Department of Social Services or the State of Connecticut. The views and opinions expressed are those of the authors. Connecticut Voices for Children 14

16 Table 1. Mothers with HUSKY A by Coverage Group Total HUSKY A Coverage Groups At Time of the Birth Total Family a Child b Pregnancyrelated c 11,374 (100.0%) 5,633 (49.5%) 320 (2.8%) 5,421 (47.7%) Maternal age: 19 and under 1,760 (15.5%) 584 (10.4%) 263 (82.2%) 913 (16.8%) 20 to 29 7,019 (61.7%) 3,383 (60.1%) 53 (16.6%) 3,583 (66.1%) 30 to 39 2,408 (21.1%) 1,556 (27.6%) 4 ( 1.3%) 848 (15.6%) 40 and over 187 ( 1.6%) 110 ( 2.0%) 0 ( 0.0%) 77 ( 1.4%) Maternal race/ethnicity: White non-hispanic 4,345 (38.2%) 1,836 (32.6%) 67 (20.9%) 2,442 (45.0%) Black non-hispanic 2,571 (22.6%) 1,344 (23.9%) 82 (25.6%) 1,145 (21.1%) Other non-hispanic 538 ( 4.7%) 247 ( 4.4%) 8 ( 2.5%) 283 ( 5.2%) Hispanic 3,920 (34.5%) 2,206 (39.2%) 163 (50.9%) 1,551 (28.6%) Primary language: English 10,480 (92.1%) 5,094 (90.4%) 285 (89.1%) 5,101 (94.1%) Non-English 894 ( 7.9%) 539 ( 9.6%) 35 (10.9%) 320 ( 5.9%) Maternal education: Less than 12 years 2,668 (23.5%) 1,579 (28.0%) 171 (53.4%) 918 (16.9%) 12 years 4,930 (43.3%) 2,419 (42.9%) 128 (40.0%) 2,383 (44.0%) More than 12 years 3,776 (33.2%) 1,635 (29.0%) 21 ( 6.6%) 2,120 (39.1%) Maternal residence: d Western Connecticut 3,599 (31.6%) 1,814 (32.2%) 110 (34.4%) 1,675 (30.9%) Northern Connecticut 4,312 (37.9%) 2,223 (39.5%) 153 (47.8%) 1,936 (35.7%) Southern Connecticut 3,463 (30.4%) 1,596 (28.3%) 57 (17.8%) 1,810 (33.4%) Birth order: First birth 4,908 (43.2%) 598 (10.6%) 285 (89.1%) 4,025 (74.2%) Second birth or more 6,466 (56.8%) 5,035 (89.4%) 35 (10.9%) 1,396 (25.8%) a Family coverage groups: HUSKY A for Families (F07), Transitional Medicaid for Families with Earnings (F03), Child Support Extension (F04) b Child coverage groups: HUSKY A for Children (F25), Medicaid for Children Under 21 (F12) c Pregnancy-related coverage groups: Medicaid for Pregnant Women (P01), Healthy Start Medicaid (P02) d Department of Social Service regional office catchment area, based on June 2012 (see The Western Region includes the offices in Stamford, Bridgeport, Danbury, Waterbury, and Torrington. The Northern Region includes New Britain, Hartford, Manchester, and Willimantic. The Southern Region includes New Haven, Norwich, and Middletown. Note: Maternal age, race/ethnicity, primary language, education, residence, and parity are all significantly associated with coverage group (P<.001). Connecticut Voices for Children 15

17 Table 2. Coverage Continuity Prior to the Birth by Coverage Group Enrolled in birth month Average months enrolled Percent continuously enrolled for 9 months Total 11, % Family coverage groups: 5, % Family coverage (F07) 5, % Transitional Medicaid (F03) % Child support extension (F04) % Child coverage groups: % HUSKY A for children (F25) % Medicaid-children <21 (F12) % Pregnancy-related coverage groups: 5, % Healthy Start (P02) 5, % Medicaid/pregnant women (P01) % Connecticut Voices for Children 16

18 Table 3. Coverage Continuity Following the Birth by Coverage Group Enrolled in birth month Average months enrolled Percent continuously enrolled for 9 months Total 11, % Family coverage groups: 5, % Family coverage (F07) 5, % Transitional Medicaid (F03) % Child support extension (F04) % Child coverage groups: % HUSKY A for children (F25) % Medicaid-children <21 (F12) % Pregnancy-related coverage groups: 5, % Healthy Start (P02) 5, % Medicaid/pregnant women (P01) % Connecticut Voices for Children 17

19 Table 4. Factors Affecting Coverage Discontinuity After the Birth Coverage Groups Race/ Ethnicity Maternal Education DSS Regional Office Maternal Language Has Other Children Factor Model parameters Significance Family groups 1.00 reference Child groups 2.63 *** Pregnancy-related 5.26 *** Black -- NS Hispanic 0.87 * White 1.00 reference Other -- NS Less than 12 years -- NS 12 years 1.00 reference More than 12 years -- NS Bridgeport 1.33 *** Danbury 1.62 *** Hartford 1.21 * Manchester -- NS Middletown -- NS New Britain -- NS New Haven 1.00 reference Norwich -- NS Stamford -- NS Torrington -- NS Waterbury -- NS Willimantic -- NS English 1.00 reference Non-English -- NS Yes 1.00 reference No -- NS Effect on Loss of Coverage in the 9 Months After the Birth } Mothers in pregnancy-related groups are over five times more likely than mothers in family groups and twice as likely as mothers in the child groups to lose coverage. }Hispanic mothers are less likely than White mothers to lose coverage. } Mothers living in Bridgeport, Danbury, or Hartford regions are more likely than mothers living in New Haven region to lose coverage. Maternal 19 and under -- NS Age reference NS 40 and over -- NS Note: In this logistic regression analysis, the coefficients represent the odds for losing coverage or experiencing a coverage gap during the nine months after the birth, as compared to the reference group, when controlling for other factors. The reference group was arbitrarily chosen and set at When the coefficient is less than 1.00, the subgroup is less likely than the reference group to have had gaps in coverage. When the coefficient is more than 1.00, the subgroup is more likely to have had gaps or lost coverage. For example, women who were in the family or child coverage groups were far less likely to lose coverage than women in the pregnancy-related coverage groups (coefficients=0.19 and 0.50, respectively). *** p <0.001 ** P <0.01 * P < 0.05 NS not statistically significant Connecticut Voices for Children 18

20 Table 5. Factors Affecting Number of Months of Continuous Coverage Coverage Groups Race/ Ethnicity Maternal Education DSS Regional Office Maternal Language Has Other Children Factor Model parameters Significance Family groups 1.95 *** Child groups 1.22 *** Pregnancy-related Reference -- Black -- NS Hispanic 0.17 * White Reference -- Other -- NS Less than 12 years -- NS 12 years Reference -- More than 12 years -- NS Bridgeport *** Danbury *** Hartford ** Manchester ** Middletown -- NS New Britain -- NS New Haven Reference -- Norwich -- NS Stamford * Torrington -- NS Waterbury -- NS Willimantic -- NS English Reference -- Non-English -- NS Yes Reference -- No -- NS Effect on Length of Coverage After the Birth }Mothers in pregnancy-related groups were covered 1-2 months less than mothers in family and child groups } Hispanic } mothers were covered about 5 days longer than White mothers Mothers living in Bridgeport, Danbury, Hartford, Manchester, and Stamford regions were covered 1-3 weeks less than mothers in New Haven region Maternal 19 and under -- NS Age Reference NS 40 and over -- NS Note: In this linear regression analysis, the coefficients represent the number of months that coverage was longer or shorter for mothers in each group, compared to the reference group when controlling for other factors. The reference group was arbitrarily chosen. When the coefficient is positive, the subgroup was covered longer than the reference group. When the coefficient is negative, the subgroup was covered for less time than the reference group. For example, women who were in the family coverage groups were covered for nearly 2 months longer on average than women in the pregnancy-related coverage groups (coefficient=1.95). *** p <0.001 ** P <0.01 * P < 0.05 NS not statistically significant Connecticut Voices for Children 19